Accepted Preprint
Biofuels and Environmental Biotechnology
Biotechnology and Bioengineering DOI 10.1002/bit.25447
Metabolomic and 13C-Metabolic Flux Analysis of a Xylose-Consuming † Saccharomyces cerevisiae Strain Expressing Xylose Isomerase Thomas M. Wasylenko and Gregory Stephanopoulos Department of Chemical Engineering, Massachusetts Institute of Technology, Cambridge, MA 02139, USA
Correspondence:
Prof. Gregory Stephanopoulos Dept. of Chemical Engineering Massachusetts Institute of Technology Room 56-469C 77 Massachusetts Ave Cambridge, MA 02139 Phone: (617) 253-4583 Fax: (617) 253-3122 Email: [email protected]
Running Title: Xylose Metabolic Flux Analysis
†
This article has been accepted for publication and undergone full peer review but has not been through the copyediting, typesetting, pagination and proofreading process, which may lead to differences between this version and the Version of Record. Please cite this article as doi: [10.1002/bit.25447]
Additional Supporting Information may be found in the online version of this article.
© 2014 Wiley Periodicals, Inc. Received May 29, 2014; Revision Received August 11, 2014; Accepted August 27, 2014
1
Accepted Preprint
Abstract Over the past two decades significant progress has been made in the engineering of
xylose-consuming Saccharomyces cerevisiae strains for production of lignocellulosic biofuels.
However, the ethanol productivities achieved on xylose are still significantly lower than those observed on glucose for reasons that are not well understood. We have undertaken an analysis of central carbon metabolite pool sizes and metabolic fluxes on glucose and on xylose under aerobic and anaerobic conditions in a strain capable of rapid xylose assimilation via xylose isomerase in order to investigate factors that may limit the rate of xylose fermentation. We find that during xylose utilization the flux through the non-oxidative PPP is high but the flux through the oxidative PPP is low, highlighting an advantage of the strain employed in this study. Furthermore, xylose fails to elicit the full carbon catabolite repression response that is characteristic of glucose fermentation in S. cerevisiae. We present indirect evidence that the incomplete activation of the fermentation program on xylose results in a bottleneck in lower glycolysis, leading to inefficient re-oxidation of NADH produced in glycolysis.
Keywords 13
C-Metabolic Flux Analysis; Carbon Catabolite Repression; Cellulosic Ethanol; Metabolomics;
Saccharomyces cerevisiae; Xylose
2
Accepted Preprint
Introduction In recent years, increasing energy demand and concerns about climate change and the
sustainability of heavy reliance on fossil fuels have motivated the development of technologies for production of liquid fuels from plant biomass. Ethanol, which serves as a fuel additive, can be readily produced by fermentation of hexose sugars derived from cornstarch and sucrose. However, production of feed stocks such as corn and sugarcane requires large amounts of arable land and may compete with the food supply. Lignocellulosic feed stocks represent an attractive
alternative, but technologies for production of liquid fuels from lignocellulosic material are relatively immature. The Baker’s yeast Saccharomyces cerevisiae is a promising biocatalyst for production of liquid fuels from lignocellulosic biomass because its high rates of ethanol production under anaerobic conditions and high ethanol tolerance allow ethanol to be produced at high yield, productivity, and final titer. S. cerevisiae also exhibits relatively high tolerance to inhibitors such as furan derivatives, weak acids, and phenolics present in lignocellulosic hydrolysates (Almeida et al., 2007; Lau et al., 2010), and the insusceptibility of yeast to bacteriophage and its ability to grow at low pH minimize the risk of contamination, allowing the avoidance of costs associated with reactor sterilization in industrial processes. However S. cerevisiae cannot natively metabolize the pentose sugars xylose and arabinose, which make up more than one-third of the carbohydrate biomass in some agricultural residues such as corn stover, wheat straw, and bagasse, with xylose being by far the more abundant of the two (van Maris et al., 2006). For production of biofuels from lignocellulosic feed stocks to be cost-
effective, it will be necessary to effect the conversion of all sugars present in hydrolysates to liquid fuels (Stephanopoulos, 2007; Carroll and Somerville, 2009). The ability to consume xylose can be conferred on S. cerevisiae strains by introduction of
a heterologous pathway for conversion of xylose to its isomer xylulose. While many bacteria use 3
Accepted Preprint
a xylose isomerase (XI) enzyme to catalyze this conversion directly without the use of pyridine nucleotide cofactors, xylose-consuming eukaryotes generally effect the isomerization through a two-step redox pathway in which xylose reductase (XR) first catalyzes the reduction of xylose to xylitol, which is then oxidized via xylitol dehydrogenase (XDH) to form xylulose. Initial attempts to express heterologous xylA (encoding XI) genes in S. cerevisiae were unsuccessful; in several cases putative xylA transcripts were detected in Northern blots but putative XI protein products were insoluble and inactive (Sarthy et al., 1987; Amore et al., 1989; Gárdonyi and Hahn-Hägerdal, 2003).
Consequently, the majority of xylose-consuming strains have been
constructed using the XR-XDH pathway. However, while XR uses NADPH as its preferred cofactor substrate, XDH is strictly NAD+-dependent. This mismatch in cofactor specificities results in a “cofactor imbalance” whereby NADP+ and NADH accumulate (and NADPH and NAD+ are depleted). The accumulation of NADH is especially problematic under industrially relevant anaerobic conditions. Without oxygen as a terminal electron acceptor, NADH cannot be efficiently re-oxidized to NAD+, severely inhibiting xylose metabolism (Bruinenberg et al., 1983; Bruinenberg et al., 1984).
In early xylose-consuming S. cerevisiae strains the low
availability of NAD+ for the XDH reaction also resulted in secretion of large amounts of the byproduct xylitol (Kötter and Ciriacy, 1993; Tantirungkij et al., 1993), compromising ethanol yield. A major breakthrough occurred with the discovery that the anaerobic fungus Piromyces
sp. strain E2 metabolizes xylose using the XI pathway (Harhangi et al., 2003). The XI from this organism was functionally expressed in S. cerevisiae (Kuyper et al., 2003), and both
evolutionary and rational metabolic engineering were used to construct efficient xyloseconsuming strains capable of anaerobic growth on xylose (Kuyper et al., 2004; Kuyper et al., 2005a; Kuyper et al., 2005b). Our lab has recently used a similar approach along with a multi-
4
Accepted Preprint
stage evolutionary strategy to engineer the S. cerevisiae strain H131-A3-ALCS, the fastest xylose-
consuming strain reported to date (Zhou et al., 2012). However, the rates of growth and ethanol production on xylose are still significantly lower than those on glucose for reasons that are not completely understood. Many hypotheses for bottlenecks in xylose metabolism have been presented. These
include xylose transport, which may be especially limiting at low extracellular xylose concentrations (Gárdonyi et al., 2003; Runquist et al., 2009a; Runquist et al., 2010; Young et al., 2012); conversion of xylose to xylulose (Lönn et al., 2003; Jeppsson et al., 2003; Karhumaa et al., 2005; Kim et al., 2012); phosphorylation of xylulose (Toivari et al., 2001; Jin et al., 2003); and conversion of xylulose-5-phosphate (X5P) to the glycolytic intermediates fructose-6-
phosphate (F6P) and glyceraldehyde-3-phosophate (GAP) via the reactions of the non-oxidative Pentose Phosphate Pathway (PPP) (Walfridsson et al., 1995; Kuyper et al., 2005a). Once xylose is metabolized to F6P and GAP xylose metabolism is in principle identical to glucose metabolism.
However while glucose induces a strong carbon catabolite repression (CCR)
response in S. cerevisiae (Gancedo, 2008), there is evidence that xylose is not recognized as a fermentable carbon source and fails to fully activate the CCR program (Jin et al., 2004; Salusjärvi et al., 2008). Consequently, bottlenecks downstream of GAP could result from altered gene expression during xylose utilization. In this study, we sought to characterize the metabolism of strain H131-A3-ALCS by
quantifying central carbon metabolite pool sizes and using
13
C-Metabolic Flux Analysis (MFA)
to estimate the fluxes through central metabolism in order to identify rate-limiting steps in xylose utilization. Although previous metabolomic and MFA studies have been conducted on xyloseconsuming S. cerevisiae strains, many of these studies investigated strains that utilized the XR-
5
Accepted Preprint
XDH pathway to effect xylose isomerization (Wahlbom et al., 2001; Pitkänen et al., 2003; Sonderegger et al., 2004; Grotkjaer et al., 2005; Klimacek et al., 2010; Feng and Zhao, 2013a; Feng and Zhao, 2013b; Matsushika et al., 2013). Consequently, in these strains the redox cofactor imbalance is expected to exert a large influence on metabolism. Moreover, many of the strains employed exhibited low xylose consumption rates and growth rates, and in several cases analysis of metabolite pool sizes revealed signs of carbon starvation (Klimacek et al., 2010; Bergdahl et al., 2012; Matsushika et al., 2013). In this study, we investigated a strain capable of rapid xylose utilization via the xylose isomerase pathway in order to observe the differences between glucose and xylose metabolism under aerobic and anaerobic conditions with high rates of xylose consumption and in the absence of the redox cofactor imbalance associated with action of the XR-XDH pathway. We present indirect evidence that in this strain there is an apparent bottleneck in xylose metabolism downstream of GAP, in the lower glycolysis pathway.
Materials and Methods
Strain and Culture Conditions All experiments were performed with a previously engineered xylose-consuming S.
cerevisiae strain similar to H131-A3-ALCS (Zhou et al., 2012). The strain was cultivated at 30 °C in minimal medium containing 6.7 g/l Yeast Nitrogen Base (YNB) without amino acids (Difco) as a source of salts, vitamins, and trace elements and 20 g/l of either glucose (YNBG) or xylose (YNBX) as the sole carbon source. Media were supplemented with 0.42 g/l Tween 80 and 0.01 g/l ergosterol to facilitate growth under anaerobic conditions (Andreasen and Stier, 1953; Andreasen and Stier, 1954).
6
Accepted Preprint
The yeast strain was cultivated in both YNBG and YNBX medium under both aerobic
and anaerobic conditions, resulting in a total of four culture conditions: Glucose Aerobic (GA), Glucose Anaerobic (GN), Xylose Aerobic (XA), and Xylose Anaerobic (XN). For each culture condition, 5 ml starter cultures were inoculated from 15% glycerol -80 °C freezer stocks. 50 ml main cultures grown in 250 ml bottles (VWR, 89000-236) were inoculated with 50 µl (0.1% by volume) of the starter culture with the same carbon source and aeration condition. All cultures were shaken at 250 rpm to facilitate mixing. Aerobic starter cultures were grown in aerobic culture tubes (BD Falcon, #352059). The
250 ml aerobic main culture bottle caps were left loose, allowing exchange between the culture headspace and the (aerobic) ambient atmosphere. Anaerobic starter cultures were grown in airtight Hungate tubes (ChemGlass, #CLS-4208-01).
Anaerobic starter culture media were
sparged with Ultra High Purity (UHP) Nitrogen (Airgas) for 3 min and left in an anaerobic chamber (Coy Labs) under an atmosphere of nitrogen and hydrogen for approximately 12 h prior to inoculation.
Anaerobic main cultures were grown with 250 ml bottle caps completely
tightened, preventing aeration from the ambient atmosphere. Anaerobic main culture media were sparged with UHP Nitrogen for 15 min and left in the anaerobic chamber for 12 h prior to inoculation. Inoculation and all sampling were conducted inside the anaerobic chamber. For 13C-MFA, the YNBG and YNBX carbon sources were 20 g/l 1,2-13C2-glucose and 20
g/l 1,2-13C2-xylose (Cambridge Isotope Laboratories), respectively. The 13C-MFA cultures were
otherwise identical.
7
Accepted Preprint
Analytical Methods Cell density was monitored by measuring the absorbance (“optical density”) of yeast
cultures at 600 nm (OD600) using an Ultrospec 2100 pro UV/Visible Spectrophotometer (Amersham Biosciences).
Cell density was calculated using pre-determined correlations
between Dry Cell Weight (DCW) and OD600. Glucose, xylose, ethanol, glycerol, and acetate
concentrations were determined by High-Performance Liquid Chromatography (HPLC). Culture samples were filtered through 0.20 µm Nylon syringe filters (Microliter Analytical #F13-20201GF) and supernatants were analyzed on an Agilent 1200 series HPLC system equipped with a refractive index detector. Analytes were separated on an Aminex HPX-87H column (Bio-Rad) with 5 mM sulfuric acid mobile phase at a flow rate of 0.6 ml/min and a temperature of 55 °C.
Estimation of Extracellular Fluxes Specific growth rates, specific sugar consumption rates, and specific ethanol, glycerol,
and acetate production rates were estimated from OD600 and HPLC data from five biological replicate cultures for each culture condition. For each biological replicate culture growth rate
was determined by fitting the OD600 data to a function of the form:
where
is the measured value of OD600 at time
and
the specific
is the fitted value of OD600 at
time zero. Parameter estimation was achieved using the Matlab function lsqnonlin. To determine specific metabolite consumption and production rates for biological
replicate culture , the concentration of each metabolite slope of the best-fit line growth rate
was plotted against OD600 and the
was computed. For each culture condition the average specific
and the average slope
were computed by averaging the
8
and
,
Accepted Preprint
respectively. Uncertainties were assumed to be equal to the standard deviations of the . The specific consumption or production rate of metabolite
where in
and
is then equal to:
is the biomass concentration that is equivalent to one OD600 unit. The relative error
was assumed to be equal to the sum of the relative errors in ,
and
.
Preparation of Uniformly 13C-Labeled Cell Extract for Internal Standard Cells were grown in medium containing 10 g/l U-13C6-glucose as the sole carbon source
and 6.7 g/l YNB without amino acids. 5 ml aerobic starter cultures were inoculated from freezer stocks (as above). 40 ml shake flask cultures were inoculated from starter cultures and harvested at OD600
1. Uniformly 13C-labeled metabolite extracts were prepared using a protocol similar
to the one described for preparation of samples for pool size measurements and
13
C-MFA (see
below). Extracts were stored at -80 °C and subsequently dried under airflow using a Pierce Reacti-Therm III Heating/Stirring Module and resuspended in 100 µl Millipore water. A total of 48 extracts (four extracts from each of 12 shake flasks) were pooled and the pooled cell extract was aliquoted. Aliquots were stored at -80 °C for future use as internal standard.
Metabolite Extractions for Pool Size Measurements and 13C-Metabolic Flux Analysis Cultures were harvested in mid-exponential phase (OD600
0.5-0.6). 7.5 ml culture was
quenched in 37.5 ml pure methanol (Canelas et al., 2008) held at low temperature (
Biofuels and Environmental Biotechnology
Biotechnology and Bioengineering DOI 10.1002/bit.25447
Metabolomic and 13C-Metabolic Flux Analysis of a Xylose-Consuming † Saccharomyces cerevisiae Strain Expressing Xylose Isomerase Thomas M. Wasylenko and Gregory Stephanopoulos Department of Chemical Engineering, Massachusetts Institute of Technology, Cambridge, MA 02139, USA
Correspondence:
Prof. Gregory Stephanopoulos Dept. of Chemical Engineering Massachusetts Institute of Technology Room 56-469C 77 Massachusetts Ave Cambridge, MA 02139 Phone: (617) 253-4583 Fax: (617) 253-3122 Email: [email protected]
Running Title: Xylose Metabolic Flux Analysis
†
This article has been accepted for publication and undergone full peer review but has not been through the copyediting, typesetting, pagination and proofreading process, which may lead to differences between this version and the Version of Record. Please cite this article as doi: [10.1002/bit.25447]
Additional Supporting Information may be found in the online version of this article.
© 2014 Wiley Periodicals, Inc. Received May 29, 2014; Revision Received August 11, 2014; Accepted August 27, 2014
1
Accepted Preprint
Abstract Over the past two decades significant progress has been made in the engineering of
xylose-consuming Saccharomyces cerevisiae strains for production of lignocellulosic biofuels.
However, the ethanol productivities achieved on xylose are still significantly lower than those observed on glucose for reasons that are not well understood. We have undertaken an analysis of central carbon metabolite pool sizes and metabolic fluxes on glucose and on xylose under aerobic and anaerobic conditions in a strain capable of rapid xylose assimilation via xylose isomerase in order to investigate factors that may limit the rate of xylose fermentation. We find that during xylose utilization the flux through the non-oxidative PPP is high but the flux through the oxidative PPP is low, highlighting an advantage of the strain employed in this study. Furthermore, xylose fails to elicit the full carbon catabolite repression response that is characteristic of glucose fermentation in S. cerevisiae. We present indirect evidence that the incomplete activation of the fermentation program on xylose results in a bottleneck in lower glycolysis, leading to inefficient re-oxidation of NADH produced in glycolysis.
Keywords 13
C-Metabolic Flux Analysis; Carbon Catabolite Repression; Cellulosic Ethanol; Metabolomics;
Saccharomyces cerevisiae; Xylose
2
Accepted Preprint
Introduction In recent years, increasing energy demand and concerns about climate change and the
sustainability of heavy reliance on fossil fuels have motivated the development of technologies for production of liquid fuels from plant biomass. Ethanol, which serves as a fuel additive, can be readily produced by fermentation of hexose sugars derived from cornstarch and sucrose. However, production of feed stocks such as corn and sugarcane requires large amounts of arable land and may compete with the food supply. Lignocellulosic feed stocks represent an attractive
alternative, but technologies for production of liquid fuels from lignocellulosic material are relatively immature. The Baker’s yeast Saccharomyces cerevisiae is a promising biocatalyst for production of liquid fuels from lignocellulosic biomass because its high rates of ethanol production under anaerobic conditions and high ethanol tolerance allow ethanol to be produced at high yield, productivity, and final titer. S. cerevisiae also exhibits relatively high tolerance to inhibitors such as furan derivatives, weak acids, and phenolics present in lignocellulosic hydrolysates (Almeida et al., 2007; Lau et al., 2010), and the insusceptibility of yeast to bacteriophage and its ability to grow at low pH minimize the risk of contamination, allowing the avoidance of costs associated with reactor sterilization in industrial processes. However S. cerevisiae cannot natively metabolize the pentose sugars xylose and arabinose, which make up more than one-third of the carbohydrate biomass in some agricultural residues such as corn stover, wheat straw, and bagasse, with xylose being by far the more abundant of the two (van Maris et al., 2006). For production of biofuels from lignocellulosic feed stocks to be cost-
effective, it will be necessary to effect the conversion of all sugars present in hydrolysates to liquid fuels (Stephanopoulos, 2007; Carroll and Somerville, 2009). The ability to consume xylose can be conferred on S. cerevisiae strains by introduction of
a heterologous pathway for conversion of xylose to its isomer xylulose. While many bacteria use 3
Accepted Preprint
a xylose isomerase (XI) enzyme to catalyze this conversion directly without the use of pyridine nucleotide cofactors, xylose-consuming eukaryotes generally effect the isomerization through a two-step redox pathway in which xylose reductase (XR) first catalyzes the reduction of xylose to xylitol, which is then oxidized via xylitol dehydrogenase (XDH) to form xylulose. Initial attempts to express heterologous xylA (encoding XI) genes in S. cerevisiae were unsuccessful; in several cases putative xylA transcripts were detected in Northern blots but putative XI protein products were insoluble and inactive (Sarthy et al., 1987; Amore et al., 1989; Gárdonyi and Hahn-Hägerdal, 2003).
Consequently, the majority of xylose-consuming strains have been
constructed using the XR-XDH pathway. However, while XR uses NADPH as its preferred cofactor substrate, XDH is strictly NAD+-dependent. This mismatch in cofactor specificities results in a “cofactor imbalance” whereby NADP+ and NADH accumulate (and NADPH and NAD+ are depleted). The accumulation of NADH is especially problematic under industrially relevant anaerobic conditions. Without oxygen as a terminal electron acceptor, NADH cannot be efficiently re-oxidized to NAD+, severely inhibiting xylose metabolism (Bruinenberg et al., 1983; Bruinenberg et al., 1984).
In early xylose-consuming S. cerevisiae strains the low
availability of NAD+ for the XDH reaction also resulted in secretion of large amounts of the byproduct xylitol (Kötter and Ciriacy, 1993; Tantirungkij et al., 1993), compromising ethanol yield. A major breakthrough occurred with the discovery that the anaerobic fungus Piromyces
sp. strain E2 metabolizes xylose using the XI pathway (Harhangi et al., 2003). The XI from this organism was functionally expressed in S. cerevisiae (Kuyper et al., 2003), and both
evolutionary and rational metabolic engineering were used to construct efficient xyloseconsuming strains capable of anaerobic growth on xylose (Kuyper et al., 2004; Kuyper et al., 2005a; Kuyper et al., 2005b). Our lab has recently used a similar approach along with a multi-
4
Accepted Preprint
stage evolutionary strategy to engineer the S. cerevisiae strain H131-A3-ALCS, the fastest xylose-
consuming strain reported to date (Zhou et al., 2012). However, the rates of growth and ethanol production on xylose are still significantly lower than those on glucose for reasons that are not completely understood. Many hypotheses for bottlenecks in xylose metabolism have been presented. These
include xylose transport, which may be especially limiting at low extracellular xylose concentrations (Gárdonyi et al., 2003; Runquist et al., 2009a; Runquist et al., 2010; Young et al., 2012); conversion of xylose to xylulose (Lönn et al., 2003; Jeppsson et al., 2003; Karhumaa et al., 2005; Kim et al., 2012); phosphorylation of xylulose (Toivari et al., 2001; Jin et al., 2003); and conversion of xylulose-5-phosphate (X5P) to the glycolytic intermediates fructose-6-
phosphate (F6P) and glyceraldehyde-3-phosophate (GAP) via the reactions of the non-oxidative Pentose Phosphate Pathway (PPP) (Walfridsson et al., 1995; Kuyper et al., 2005a). Once xylose is metabolized to F6P and GAP xylose metabolism is in principle identical to glucose metabolism.
However while glucose induces a strong carbon catabolite repression (CCR)
response in S. cerevisiae (Gancedo, 2008), there is evidence that xylose is not recognized as a fermentable carbon source and fails to fully activate the CCR program (Jin et al., 2004; Salusjärvi et al., 2008). Consequently, bottlenecks downstream of GAP could result from altered gene expression during xylose utilization. In this study, we sought to characterize the metabolism of strain H131-A3-ALCS by
quantifying central carbon metabolite pool sizes and using
13
C-Metabolic Flux Analysis (MFA)
to estimate the fluxes through central metabolism in order to identify rate-limiting steps in xylose utilization. Although previous metabolomic and MFA studies have been conducted on xyloseconsuming S. cerevisiae strains, many of these studies investigated strains that utilized the XR-
5
Accepted Preprint
XDH pathway to effect xylose isomerization (Wahlbom et al., 2001; Pitkänen et al., 2003; Sonderegger et al., 2004; Grotkjaer et al., 2005; Klimacek et al., 2010; Feng and Zhao, 2013a; Feng and Zhao, 2013b; Matsushika et al., 2013). Consequently, in these strains the redox cofactor imbalance is expected to exert a large influence on metabolism. Moreover, many of the strains employed exhibited low xylose consumption rates and growth rates, and in several cases analysis of metabolite pool sizes revealed signs of carbon starvation (Klimacek et al., 2010; Bergdahl et al., 2012; Matsushika et al., 2013). In this study, we investigated a strain capable of rapid xylose utilization via the xylose isomerase pathway in order to observe the differences between glucose and xylose metabolism under aerobic and anaerobic conditions with high rates of xylose consumption and in the absence of the redox cofactor imbalance associated with action of the XR-XDH pathway. We present indirect evidence that in this strain there is an apparent bottleneck in xylose metabolism downstream of GAP, in the lower glycolysis pathway.
Materials and Methods
Strain and Culture Conditions All experiments were performed with a previously engineered xylose-consuming S.
cerevisiae strain similar to H131-A3-ALCS (Zhou et al., 2012). The strain was cultivated at 30 °C in minimal medium containing 6.7 g/l Yeast Nitrogen Base (YNB) without amino acids (Difco) as a source of salts, vitamins, and trace elements and 20 g/l of either glucose (YNBG) or xylose (YNBX) as the sole carbon source. Media were supplemented with 0.42 g/l Tween 80 and 0.01 g/l ergosterol to facilitate growth under anaerobic conditions (Andreasen and Stier, 1953; Andreasen and Stier, 1954).
6
Accepted Preprint
The yeast strain was cultivated in both YNBG and YNBX medium under both aerobic
and anaerobic conditions, resulting in a total of four culture conditions: Glucose Aerobic (GA), Glucose Anaerobic (GN), Xylose Aerobic (XA), and Xylose Anaerobic (XN). For each culture condition, 5 ml starter cultures were inoculated from 15% glycerol -80 °C freezer stocks. 50 ml main cultures grown in 250 ml bottles (VWR, 89000-236) were inoculated with 50 µl (0.1% by volume) of the starter culture with the same carbon source and aeration condition. All cultures were shaken at 250 rpm to facilitate mixing. Aerobic starter cultures were grown in aerobic culture tubes (BD Falcon, #352059). The
250 ml aerobic main culture bottle caps were left loose, allowing exchange between the culture headspace and the (aerobic) ambient atmosphere. Anaerobic starter cultures were grown in airtight Hungate tubes (ChemGlass, #CLS-4208-01).
Anaerobic starter culture media were
sparged with Ultra High Purity (UHP) Nitrogen (Airgas) for 3 min and left in an anaerobic chamber (Coy Labs) under an atmosphere of nitrogen and hydrogen for approximately 12 h prior to inoculation.
Anaerobic main cultures were grown with 250 ml bottle caps completely
tightened, preventing aeration from the ambient atmosphere. Anaerobic main culture media were sparged with UHP Nitrogen for 15 min and left in the anaerobic chamber for 12 h prior to inoculation. Inoculation and all sampling were conducted inside the anaerobic chamber. For 13C-MFA, the YNBG and YNBX carbon sources were 20 g/l 1,2-13C2-glucose and 20
g/l 1,2-13C2-xylose (Cambridge Isotope Laboratories), respectively. The 13C-MFA cultures were
otherwise identical.
7
Accepted Preprint
Analytical Methods Cell density was monitored by measuring the absorbance (“optical density”) of yeast
cultures at 600 nm (OD600) using an Ultrospec 2100 pro UV/Visible Spectrophotometer (Amersham Biosciences).
Cell density was calculated using pre-determined correlations
between Dry Cell Weight (DCW) and OD600. Glucose, xylose, ethanol, glycerol, and acetate
concentrations were determined by High-Performance Liquid Chromatography (HPLC). Culture samples were filtered through 0.20 µm Nylon syringe filters (Microliter Analytical #F13-20201GF) and supernatants were analyzed on an Agilent 1200 series HPLC system equipped with a refractive index detector. Analytes were separated on an Aminex HPX-87H column (Bio-Rad) with 5 mM sulfuric acid mobile phase at a flow rate of 0.6 ml/min and a temperature of 55 °C.
Estimation of Extracellular Fluxes Specific growth rates, specific sugar consumption rates, and specific ethanol, glycerol,
and acetate production rates were estimated from OD600 and HPLC data from five biological replicate cultures for each culture condition. For each biological replicate culture growth rate
was determined by fitting the OD600 data to a function of the form:
where
is the measured value of OD600 at time
and
the specific
is the fitted value of OD600 at
time zero. Parameter estimation was achieved using the Matlab function lsqnonlin. To determine specific metabolite consumption and production rates for biological
replicate culture , the concentration of each metabolite slope of the best-fit line growth rate
was plotted against OD600 and the
was computed. For each culture condition the average specific
and the average slope
were computed by averaging the
8
and
,
Accepted Preprint
respectively. Uncertainties were assumed to be equal to the standard deviations of the . The specific consumption or production rate of metabolite
where in
and
is then equal to:
is the biomass concentration that is equivalent to one OD600 unit. The relative error
was assumed to be equal to the sum of the relative errors in ,
and
.
Preparation of Uniformly 13C-Labeled Cell Extract for Internal Standard Cells were grown in medium containing 10 g/l U-13C6-glucose as the sole carbon source
and 6.7 g/l YNB without amino acids. 5 ml aerobic starter cultures were inoculated from freezer stocks (as above). 40 ml shake flask cultures were inoculated from starter cultures and harvested at OD600
1. Uniformly 13C-labeled metabolite extracts were prepared using a protocol similar
to the one described for preparation of samples for pool size measurements and
13
C-MFA (see
below). Extracts were stored at -80 °C and subsequently dried under airflow using a Pierce Reacti-Therm III Heating/Stirring Module and resuspended in 100 µl Millipore water. A total of 48 extracts (four extracts from each of 12 shake flasks) were pooled and the pooled cell extract was aliquoted. Aliquots were stored at -80 °C for future use as internal standard.
Metabolite Extractions for Pool Size Measurements and 13C-Metabolic Flux Analysis Cultures were harvested in mid-exponential phase (OD600
0.5-0.6). 7.5 ml culture was
quenched in 37.5 ml pure methanol (Canelas et al., 2008) held at low temperature (
